484 



C. Barus — Fluid Volume and its 



To measure pressure, I used a Bourdon gauge,* graduated in 

 steps of 10 atm. each as far as 1000 atm. So graduated the 

 gauge registered smaller differences of pressure than could be 

 measured in terms of the volume changes of the liquid in the 

 capillary tubes. 



To test the gauge I compared it with a similar gauge gradu- 

 ated as far as 300 atm., and found the difference no larger than 

 the error of reading. Again 1 compared the guage indications 

 with the changes of the electrical resistance of mercury under 

 identical conditions of pressure, and found the results equally 

 satisfactory, as will be explained elsewhere. 



Curious results were obtained on comparing the gauge with 

 a closed air manometer. The results calculated with reference 

 to Boyle's law, i. e. for isothermal expansion, were found 

 wholly discrepant. The difficulty was interpreted by comput- 

 ing with reference to Poisson's law. Some data are given in 

 table 1, where p, t>, &, denote pressure, volume and the specific 

 heat ratio, respectively. The error at p — 100 is due to time 

 lost in making the measurement, since it is here that the 

 volume changes are large and require specially adjusted tele- 

 scopes for close reading. 



Table 1. — Compressibility of air. Test of Gauge. 



p 



V 



lg v k p 





P 



V 



\gy h p 



aim. 









atm. 







100 



1030 



6-25 





100 



1040 



625 



200 



550 



6-16 





200 



552 



6-17 



300 



400 



6-15 





300 



393 



6-14 



400 



329 



6-15 





400 



324 



6-16 



500 











500 



285 



6-16 



Quick manipulation presupposed, this method is probably 

 available for the measurement of Jc, particularly with reference 

 to its pressure variation. It follows also that when work is done 

 in capillary tubes, much time must be allowed before the 

 observed volume changes can be considered isothermal. 



Since the gauge is based on Hooke's law, errors of the zero 

 point are eliminated by working between indicated pressure 

 and taking differences of reading. So far as the relative pres- 

 sures are concerned, it is then merely necessary that the at- 

 tached dial and mechanism be virtually a scale of equal parts. 



* I think Mr. Tait who has constructed other gauges based on Hooke's law, 

 under-estimates the ingenuity of Bourdon's gauge. By increasing the number of 

 coils, or by suitably adjusting a mirror index, the gauge can be made serviceable 

 without a multiplying gear. Read off directly the only limit to the scientific 

 efficiency of such a gauge is the viscosity of the Bourdon tube. My own experi- 

 ments on improving this gauge have thus far been failures, chiefly however 

 because of insufficient facilities for working with steel tube. 



